Canine gene microarray technology has recently become available. This technology uses a blood sample to detect genes that are being expressed during the course of an illness. This study will investigate whether differences in patterns of gene expression correlate with the severity of illness during sepsis and differentiate between survivors and non-survivors. The study is designed to examine the genetic mechanisms of sepsis. By using four different bacterial levels (0, 6, 12, 18 x 109 CFUs), we will produce a range of severity of infection which should produce distinct patterns of gene expression. This design allows for the detection of trends in both activation and inactivation of the genetic pathways involved in producing septic pathophysiology. In addition, unlike small animal models, the canine sepsis model allows for serial measurements over time and will allow for comparison of genetic changes in survivors vs. non-survivors. This information could identify new targets for therapy, new tests that will allow for the determination of the presence and severity of sepsis, and possibly markers that predict survival